A jury concluded that an insurance agency overcharged the City of Rosemont for its services and kicked back part of the excess to Nick Boscarino, who helped the agency secure the business. Boscarino also helped Ralph Aulenta, one of the agency’s managers, hide money that Au-lenta had taken from the till. To top it off, Boscarino failed to report as income much of the ill-gotten gains and committed other tax offenses. Aulenta pleaded guilty and testified against Boscarino, who was convicted of mail fraud, money laundering, and tax crimes. His sentence is 36 months’ imprisonment, 24 months’ supervised release, a $55,000 fine, restitution of $288,670, special assessments of $1,700, and the costs of prosecution, which the judge set at $4,692 — for Boscarino is the rare criminal defendant who has legitimate assets sufficient to cover all of these monetary exactions.

Boscarino’s appellate lawyer has pursued almost every contention that trial counsel raised and lost. The result is that none of the issues has been developed in depth, and strong contentions (if any) have been buried under anemic ones. “Experienced advocates since time beyond memory have emphasized the importance of winnowing out weaker arguments on appeal and focusing on one central issue if possible, or at most on a few key issues.” Jones v. Barnes, 463 U.S. 745, 751-52, 103 S.Ct. 3308, 77 L.Ed.2d 987 (1983). We discuss only three of the contentions; the rest have been considered but are too feeble to call for exposition.

Every year that Rosemont placed its insurance through a brokerage that the parties call ABI/Acordia, Aulenta caused the firm to write a check to a corporation that Boscarino controlled. Though the money supposedly was a referral fee to compensate Boscarino for his assistance in persuading Rosemont to give ABI/Acordia the business, the check was never made out to Boscarino. He did not deposit the funds into the corporate accounts; instead he endorsed the checks to Aulenta, who returned half of the amount in monthly dollops over the coming year and kept the rest. The prosecution’s theory was that[*636] Aulenta was stealing money from ABI/Acordia and sharing half of the takings with Boscarino, in part for his assistance in disguising the transaction; the brokerage did not miss the money because Aulenta simultaneously was overbilling Rosemont, so that ABI/Aeordia’s books balanced. A jury was entitled to find that Boscarino, who has considerable experience in business, recognized that these transactions had the hallmarks of fraud rather than above-board referral fees. Corporate insiders don’t keep half of bona fide referral fees, nor are such fees paid from an insider’s personal account after such a roundabout transaction.

Because many of the payments passed through the mails, the indictment included a charge of mail fraud. 18 U.S.C. § 1341. And because Aulenta owed ABI/Acordia a fiduciary duty of loyalty, the indictment alleged that one aspect of the scheme was to defraud ABI/Acordia of Aulenta’s honest services. 18 U.S.C. § 1346. This sets up Boscarino’s challenge to his conviction for money laundering, in violation of 18 U.S.C. § 1956. Section 1956 makes it a crime to engage in financial transactions with the proceeds of “specified unlawful activity.” That phrase, a defined term, includes “any act or activity constituting an offense listed in section 1961(1) of this title”. 18 U.S.C. § 1956(c)(7)(A). Section 1961(1) in turn refers to mail fraud, in violation of § 1341, but does not mention § 1346. Because the mail fraud charge in this case included a reference to § 1346, Boscarino contends, it cannot serve as a predicate offense for money laundering — at least not unless the jury was instructed to disregard the honest-services aspect of the scheme, and his jury was not so instructed.

Whether a mail-fraud scheme that was carried out, in part, by depriving one person of another’s honest services may be a predicate offense for a money-laundering conviction is a question of first impression among the courts of appeals. But the answer is not difficult. Section 1346 does not create a separate crime. It is a definitional clause, reading in full: “For the purposes of this chapter, the term ‘scheme or artifice to defraud’ includes a scheme or artifice to deprive another of the intangible right of honest services.” The scheme to defraud itself violates § 1341, which is a listed predicate offense for the money-laundering statute.

Boscarino observes that only “proceeds” can be laundered, and depriving one’s employer of honest services need not yield “proceeds.” That’s true enough, but when the offense does create proceeds, which are laundered to hide detection, it is sensible to treat them the same as any other proceeds of mail or wire fraud. Consider, for example, the bribery of public officials, as in United States v. Murphy, 768 F.2d 1518 (7th Cir.1985). Judge Murphy took money from litigants in cases over which he presided. Doing this deprived the public of his honest services. He did not take money from the public coffers, but the bribes were “proceeds” of the scheme to defraud, and if he had engaged in financial transactions with these proceeds Judge Murphy could have been convicted of money laundering as well as the scheme to defraud the public. Just so here. Aulenta deprived ABI/Acordia of both his honest services and the firm’s money; the cash, which he shared with Boscarino, was “proceeds” that the two could (and did) launder to disguise the money’s origin.

Boscarino’s jury was instructed that it could convict him of the § 1956 charge only if it found that he engaged in financial transactions with the “proceeds” of a fraud. There is no chance that the jury thought that Boscarino laundered Aulen-ta’s chicanery. One launders money (or[*637] clothes) but not “services,” honest or otherwise. Anyway, Boscarino did not ask for an instruction that would have made it pellucid that “proceeds” means money and other things of value to third parties, rather than Aulenta’s duty of loyalty.

Restitution is the second subject we must cover. The district court ordered Boscarino to repay ABI/Acordia what Aulenta had extracted during the scheme. That’s inappropriate, Boscarino contends, because Rosemont rather than ABI/Acordia is the victim; after all, Aulenta obtained that money for the brokerage in the first place by bilking the City. One response is that even a thief can be the victim of a crime. See, e.g., Levin v. United States, 338 F.2d 265 (D.C.Cir.1965). ABI/Acordia was not entitled to this money vis-a-vis Rosemont, but it has rights superior to those of Aulenta and Boscarino. See, e.g., Anderson v. Gouldberg, 51 Minn. 294, 296, 53 N.W. 636, 637 (1892); Ward v. People, 3 Hill 395 (N.Y.1842). Another, and more functional, response is that ABI/Acordia is just a way station for the funds. Once Boscarino reimburses the immediate victim, ABI/Acordia will be able to repay Rosemont. Instead of determining the ultimate incidence of costs created by criminal activity, judges should direct restitution to the immediate victim; other persons’ rights in the funds then may be sorted out under normal rules of contract and property law. See United States v. Shepard, 269 F.3d 884, 886-87 (7th Cir.2002).

Finally, we consider Boscarino’s contention that his sentence is unreasonably high. Thirty-six months falls within a properly constructed range under the Sentencing Guidelines. (For the loss involved, the range is 33 to 41 months.) Instead of comparing his sentence to the range, however, Boscarino wants us to compare it to Aulenta’s sentence. Had Aulenta not pleaded guilty, his range would have been 41-51 months. His guilty plea cut the range to 30-37 months. Because Aulenta assisted the prosecution by testifying against Boscarino, the United States proposed a reduction under U.S.S.G. § 5K1.1, and Aulenta’s actual sentence was 20 months. District judges are supposed to reduce disparity in sentencing, see 18 U.S.C. § 3553(a)(6), and Boscarino contends that it is unacceptably disparate to give the lower sentence to the more culpable offender.

Until recently we refused to address arguments by criminal defendants who sought below-Guideline sentences, at least when district judges recognized their authority to depart. See United States v. Franz, 886 F.2d 973 (7th Cir.1989). United States v. Booker, 543 U.S. 220, 125 S.Ct. 738, 160 L.Ed.2d 621 (2005), which abolished “departures” by making the Guidelines advisory, abolished this rule too. See United States v. Vaughn, 433 F.3d 917 (7th Cir.2006); United States v. Arnaout, 431 F.3d 994 (7th Cir.2005). We held in Franz that a request for a below-guideline sentence did not fit any of the categories in 18 U.S.C. § 3742(a), which authorizes appellate review of sentences at defendants’ behest. After Booker, however, an “unreasonable” sentence is an unlawful sentence, and § 3742(a)(1) authorizes the correction of any illegal sentence. Because sentences within the Guideline range are presumptively but not conclusively reasonable, we are authorized to entertain contentions that a particular Guideline sentence is unreasonably high.

This is as far as Boscarino gets, however. His argument misunderstands what § 3553(a)(6) means when saying that district judges must consider “the need to avoid unwarranted sentence disparities among defendants with similar records who have been found guilty of similar con[*638] duct”. Boscarino and Aulenta had similarly clean records before these convictions, and they engaged in similar conduct, but a sentencing difference is not a forbidden “disparity” if it is justified by legitimate considerations, such as rewards for cooperation. Indeed, before Booker a district judge was forbidden to reduce one defendant’s sentence because of a discount properly given to another. See United States v. Meza, 127 F.3d 545 (7th Cir.1997). See also, e.g., United States v. Joyner, 924 F.2d 454 (2d Cir.1991). The norms of sentencing are no longer so unyielding; Booker turns rules into standards, and the rule of Meza is one of those that have been transfigured. See United States v. Newsom, 428 F.3d 685, 688-89 (7th Cir.2005).

Still, Booker is about the allocation of fact-finding authority between judge and jury, and about the burden of persuasion. It does not change rules of law. See, e.g., United States v. Duncan, 413 F.3d 680, 683 (7th Cir.2005); United States v. Rivera, 411 F.3d 864, 866-67 (7th Cir.2005); United States v. Lee, 399 F.3d 864, 866 (7th Cir.2005); McReynolds v. United States, 397 F.3d 479, 481 (7th Cir.2005). A reason bad before Booker (e.g., alienage, race, sex) is bad today. One rule of law that preceded Booker, and retains vitality after it, is that a sentencing difference based on one culprit’s assistance to the prosecution is legally appropriate.

There would be considerably less cooperation — and thus more crime — if those who assist prosecutors could not receive lower sentences compared to those who fight to the last. Neither Booker nor § 3553(a)(6) removes the incentive for cooperation — and because this incentive takes the form of a lower sentence for a cooperator than for an otherwise-identical defendant who does not cooperate, the reduction cannot be illegitimate. After all, § 3553(a)(6) disallows “unwarranted sentence disparities” (emphasis added), not all sentence differences.

Another way to put this point is to observe that the kind of “disparity” with which § 3553(a)(6) is concerned is an unjustified difference across judges (or districts) rather than among defendants to a single case. If the national norm for first offenders who gain $275,000 or so by fraud is a sentence in the range of 33 to 41 months, then system-wide sentencing disparity will increase if Boscarino’s sentence is reduced so that it comes closer to Aulen-ta’s. Instead of one low sentence, there will b.e two low sentences. But why should one culprit receive a lower sentence than some otherwise-similar offender, just because the first is “lucky” enough to have a confederate turn state’s evidence? Yet that is Boscarino’s position, which has neither law nor logic to commend it.

Sentencing disparities are at their ebb when the Guidelines are followed, for the ranges are themselves designed to treat similar offenders similarly. That was the main goal of the Sentencing Reform Act. The more out-of-range sentences that judges impose after Booker, the more disparity there will be. A sentence within a properly ascertained range therefore cannot be treated as unreasonable by reference to § 3553(a)(6).

Affirmed